MLLM-Microscope: Unlocking Hidden Structure Within Multimodal Large Language Models 事件

MLLM-Microscope: Unlocking Hidden Structure Within Multimodal Large Language Models arXiv:2606.00909v1 Announce Type: new Abstract: This work presents MLLM-Microscope, a novel system designed for analyzing the hidden representations within Multimodal Large Language Models (MLLMs). Our system evaluates the linearity, intrinsic dimension, and anisotropy of multimodal token embeddings across transformer layers. Utilizing the ScienceQA dataset, we evaluate two state-of-the-art MLLMs, LLaVA-NeXT and